In adult acute non-lymphocytic leukemia (ANLL), 30-50% of cases contain an activated N-ras oncogene or an activated Ki-ras oncogene with point mutations at codons 12, 13 or 61. These point mutations, although only a step in leukemic progression, represent useful genetic markers for distinguishing the leukemic cells from the normal cells in patients with acute leukemia. This is par- ticularly important since no suitable morphological, histochemical or immunological ANLL-specific markers are available. The long-term objective of the proposed research is to exploit the presence of an activated ras to monitor minimal residual disease in children undergoing chemotherapy for ANLL. The proposed approach includes development of procedures that utilize the coupling of the polymerase chain reaction with a forced cloning into plasmid libraries in order to increase the limits of sensitivity of detection to at least one malignant cell in 104 normal cells. Other approaches include: transfection and direct genomic sequencing to determine the frequency of childhood ANLL with an activated ras the frequency in which the ras alteration is preexistent or is a late addition to the leukemic phenotype; analysis of the extent of minimal residual disease during and after chemotherapy in ANLL patients with an activated ras oncogene and to determine whether this technique is useful for early diagnosis of relapse.